Toxic chemicals and childhood cancer: A review of the evidence Tami Gouveia-Vigeant, MPH, MSW and Joel Tickner, ScD With contributions from Richard Clapp, DSc 1 May, 2003 A Publication of the Lowell Center for Sustainable Production University of Massachusetts Lowell One University Avenue Lowell, MA 01854 978-934-2981 sustainableproduction.org 1 The Lowell Center for Sustainable Production The Lowell Center for Sustainable Production develops, studies, and promotes environmentally sound systems of production, healthy work environments, and economically viable work organizations. The Center operates on the premise that environmental quality, safe and healthy workplaces, and social accountability can be achieved while at the same time enhancing the economic life of firms. This is accomplished by broadening the fundamental design criteria for all productive activities to include an explicit and comprehensive commitment to sustainability. The Center is composed of faculty and staff at the University of Massachusetts Lowell who work directly with industrial firms, social service institutions, citizen organizations, and government agencies to promote sustainable production. © Lowell Center for Sustainable Production, University of Massachusetts Lowell 1 Department of Environmental Health, Boston University School of Public Health 1 EXECUTIVE SUMMARY Childhood cancer is the second largest cause of death to children ages 0-15 in the United States (second only to accidents), and more than 8,000 cases are diagnosed each year. In Massachusetts from 1990-1999, approximately 2,688 children ages 0-19 were diagnosed with cancer and 394 died. The overall rate of childhood cancer in Massachusetts is slightly higher than the national average—16.7 new cases versus 16.1 per 100,000 per year. African American and Latino children in Massachusetts had approximately 25% more diagnosed cancers than white and Asian and Pacific Islander children. Although childhood cancer is a relatively rare disease, cancer rates increased nearly 21% between 1975 and 1998—approximately 1% each year. Some causes of cancer can be attributed to genetic predisposition, while it is highly likely that environmental exposures, including toxic substances in our environment, food, water, and consumer products, play a role. A panel of experts convened by Mt. Sinai Hospital recently concluded that genetic predisposition accounts for no more than 20% of all childhood cancers and that the environmental attributable fraction of childhood cancer could be between 5% and 90%, depending on the type of cancer. This means that a potentially large percentage of childhood cancers is preventable. There are some well-established links between environmental exposures and childhood cancer, including: pharmaceuticals such as diethylstilbestrol (DES), an estrogen prescribed from the late 1940s to the early 1970s to prevent miscarriage; ionizing radiation; and chemotherapeutic agents. However, evidence increasingly indicates that parental and childhood exposures to certain toxic chemicals including solvents, pesticides, petrochemicals and certain industrial by-products (dioxins and polycyclic aromatic hydrocarbons) can result in childhood cancer. This report, commissioned by the Massachusetts Alliance for a Healthy Tomorrow, examines the evidence linking exposures to solvents, pesticides, petrochemicals, and certain industrial by- products with cancer in children. The report is based on examination of the published literature 2 on epidemiologic studies, animal toxicologic data, reviews of published studies and analyses of studies, case reports, fact sheets, and conference summaries. Our analysis found the following: • Epidemiologic studies have consistently found an increased likelihood of certain types of childhood cancer following parental and childhood exposure to pesticides and solvents. Studies indicate that parental exposure to certain petroleum-based chemicals and parental and childhood exposure to combustion by-products, such as dioxins and polycyclic aromatic hydrocarbons, may increase the likelihood of childhood leukemia and brain and central nervous system cancers. In one study of pesticide exposures, children with leukemia were 4 to 7 times as likely to have been exposed to pesticides used in the yard or garden compared to children without the disease. Another study found that children with leukemia were 11 times as likely to have mothers who were exposed to pesticide sprays or foggers during pregnancy compared to healthy children. Compared to children of unexposed fathers, children whose fathers were occupationally exposed to benzene and alcohols used in industrial products were nearly 6 times as likely to develop leukemia if the exposure occurred prior to the pregnancy. In Dover Township, New Jersey, researchers found that children with leukemia were 5.4 times as likely as children without leukemia to have drunk water from private wells in groundwater areas with a history of contamination from the Reich Farm Superfund site or wastewater from a nearby industrial facility. In another study, children with acute non-lymphocytic leukemia (ANLL) were 2.4 times as likely as those without ANLL to have parents who were exposed to petroleum products in their jobs. This evidence is supported by laboratory experiments and data on adult cancers from similar exposures. In most cases, the studies do not provide evidence of cancer from exposure to particular chemicals but rather mixtures or classes of chemicals (e.g., pesticides, solvents, hydrocarbons). 3 • Exposures that occur prior to conception, in the womb, and in early childhood can increase the likelihood of childhood cancer. Cancer may develop in the fetus if the germ cells (sperm and eggs) of the mother or father are damaged prior to pregnancy. Also, a fetus may be exposed to potentially harmful chemicals in utero. In such cases, the toxic substance can cross the placenta and enter the body of a developing fetus, potentially leading to cancer. Based on the literature, the types of exposures that have the strongest apparent links to childhood cancer include: parental exposure to pesticides from occupational, agricultural, home, and garden uses; parental exposure to solvents in manufacturing and painting; parental occupational exposure to hydrocarbons; maternal exposure to water contaminated with solvents; direct childhood exposure to pesticides from home and garden use; childhood exposure to solvents in drinking water; and childhood exposure to dioxins. • The evidence supporting the connection between exposure to these toxicants and childhood cancer is strongest for leukemia, brain and central nervous system cancers. It is difficult to determine the exact magnitude of the contribution of toxic chemicals to the overall burden of childhood cancer. Because the majority of chemicals in commerce—some of which are widely used in everyday products—have not been studied for their potential to cause cancer, we do not have a complete picture of the potential chemical causes of cancer in children. The links with childhood cancer have been adequately studied for only a few chemicals. Mixtures of chemicals mimicking the complex exposures that occur in everyday life have been studied even less. Since people are exposed to many chemicals and other agents simultaneously, and cancer is a rare disease, it is very difficult to establish causal links. Because of these difficulties and the costs of studies, relatively few epidemiologic studies examining the links have been conducted. Further, many studies that have been conducted have serious limitations and could be expected to provide only weak evidence about causes and childhood cancer. The lack of proof of direct 4 causal links between toxics and childhood cancer should not be construed as proof of safety. There are far more chemicals in circulation with little or no evidence of harm or safety than there are chemicals tested regularly and shown to be safe. The evidence presented in this report indicates that preventing parental and childhood exposure to chemicals suspected of causing cancer can have important health benefits. The types of chemicals examined in this report are of concern not only for their ability to cause cancer but other health effects as well—neurological and developmental harms to the fetus, for example. Preventing exposure to chemicals suspected of causing cancer is possible, as recent European policies demonstrate. The European Union will soon require that all chemicals in commercial circulation receive basic testing, and that those that are known or probable carcinogens, mutagens, or reproductive toxicants be used only when there are no safer economically and technically feasible alternatives. This common sense approach to chemical safety is likely to result in significant reductions in childhood exposure to potentially dangerous chemicals. 5 INTRODUCTION The Massachusetts Alliance for a Healthy Tomorrow asked the Lowell Center for Sustainable Production to examine the documented links between environmental toxins and cancer in children. This report is based on an examination of the published literature on epidemiologic studies, animal toxicologic data, reviews of published studies and analyses of studies, case reports, fact sheets, and conference summaries. We examine the strength of the evidence on whether exposures to pesticides, solvents, petrochemicals and combustion by-products increase the likelihood of childhood cancer. We focus particularly on leukemia and brain cancer, because they are more common compared to other cancers, and therefore studied more often. During the last two decades, concerns about the links between environmental factors, including exposure to toxic substances, and childhood cancer have increased. While there is still some debate about the exact magnitude and importance of the observed increases in childhood cancer rates over the last two decades and the causes of the increase, a growing body of evidence from laboratory studies and human epidemiologic studies suggests that toxic substances cannot be ruled out as contributors to childhood cancer. In this report, we examine the body of evidence on the relationship between toxic substance exposures and certain childhood cancers. This report reviews the evidence for certain chemical exposures for which there is increasing evidence of potential carcinogenicity in children. These chemicals include pesticides, industrial solvents, and some combustion by-products (such as dioxins) and hydrocarbons (petroleum products). We examine the evidence for each class of substance and discuss the strengths and limitations of the literature. We conclude that there is sufficient human and laboratory evidence that exposure to some common environmental chemicals can result in childhood cancer. Instituting measures to reduce parental and childhood exposures to these and other substances suspected of causing cancer, including development of safer substitutes, should play an important role in a cancer prevention strategy. 6 Cancer is most common fatal disease in children Cancer is the most common fatal disease in U.S. children, (second only to accidents among all causes), resulting in approximately 1,500 deaths per year (Zahm and Devesa, 1995). Although cancer mortality has decreased over the years due to improved detection and treatment, more than 8,000 cancer diagnoses are made in U.S. children under the age of 15 annually. Leukemia and cancers of the central nervous system (CNS), including the brain, account for approximately 50% of cancers in children, with diagnosis of leukemia and CNS cancers typically made in children under the age of 2 and 5 respectively (Zahm and Devesa, 1995; Robison, et al., 1995; Carroquino, et al., 1998; Grufferman, 1998; Schmidt, 1998). According to a 2003 U.S. Environmental Protection Agency (U.S. EPA) report, leukemia incidence increased from 24 cases per 1,000,000 children during the 1974-1978 reporting period to 28 cases per 1,000,000 children during the 1994-1998 reporting period. 1,2 The incidence of CNS tumors increased from 22 per 1,000,000 children during 1979-1983 and peaked at 30 cases per 1,000,000 children by 1993. Fortunately, incidence of CNS tumors has decreased. However, 27 out of every 1,000,000 children were diagnosed with CNS tumors, including brain tumors, between 1994 and 1998 (U.S. EPA, 2003). Overall, childhood cancer incidence rates in Massachusetts are slightly higher (about 4%, 16.7 versus 16.1 per 100,000) than the national rates which come from the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) program. The Massachusetts rate for leukemia was slightly lower, for lymphoma the rate was slightly higher, and for brain and CNS cancers they were the same as the national rate. Total childhood cancer incidence for females from 1990-1999 went up 1.6% per year, while for males it went down an average of 0.7% per year. For males and females combined the total childhood cancer incidence from 1990-1999 increased approximately 0.5% per year. Childhood cancer death rates are decreasing slightly in the state, though nearly 394 children died from cancer in Massachusetts between the years 1990 and 1999 (MDPH, 2003). 1 Incidence rate refers to the number of new cases out of a total given population in a given time period. 2 U.S. EPA data was computed for children under the age of 20 at time of diagnosis. 7 From 1995-1999, childhood cancer incidence among Latino and African-American children was approximately 25% higher (20 per 100,000) than that among white and Asian and Pacific Islander children (15 per 100,000) and childhood cancer mortality during the years 1990-1999 among African-American children was approximately 25% higher than that among white, Latino, and Asian and Pacific Islander children (MDPH, 2003). The incidence of all cancers in children in the U.S. increased nearly 21% between 1975 and 1998—approximately 1% every year for the last two decades (Zahm and Devesa, 1995; Colt and Blair, 1998; Schmidt, 1998). Some cancer researchers argue that improved technology, detection methods, and diagnoses (i.e., computerized axial tomography scans and magnetic resonance imaging) account for the rise, while others argue that if this were the case, one would expect to see cancer incidence rates flattening, which has not yet occurred (Schmidt, 1998; Kaiser, 1999). Others argue that it is impossible to miss brain cancer and leukemia because the symptoms are so painfully obvious (brain cancer) and the tests accurate (leukemia) (Kaiser, 1999). Given the increasing trend in childhood cancer incidence, and the lack of definitive explanations for it, it is important to consider the evidence for environmental chemical causes. While some researchers postulate that genes and viruses are the main contributors to any observed increase in childhood cancer, other researchers argue that genes, individual susceptibility and the environment are likely to interact in such a way as to disrupt normal cell function, leading to cancer (Zahm and Ward, 1998; Robison, et al., 1995; Carroquino, et al, 1998; Shannon, 1998; Czene, et al., 2002). A panel of experts convened by Mt. Sinai Hospital concluded that no more than 10%-20% of childhood cancer cases could be attributed to genetic predisposition; non-genetic factors, defined broadly, thus contribute to the other 80%-90%. Given that the specific causes of childhood cancer are largely unknown due to limited study, the panel concluded that the environmental attributable fraction of childhood cancer due to toxic chemical exposures was at least 5-10% and less than 80-90% (Landrigan, et al., 2002). 8 This means that there are between 400 and 7,200 new cases of childhood cancer per year in the U.S. potentially due to chemical exposures. The Mt. Sinai panel estimated that the annual cost of environmentally related childhood cancer—due to hospitalization and treatment, treatment of secondary cancers, lost parental wages, and decreased IQ due to cancer treatments—ranges from $132 million to $663 million (Landrigan, et al., 2002). Table 1. Number of Cancer Cases and Deaths by Site in Massachusetts Children Younger than 20 Years (1990-1999) 3 Cancer or Tumor Site Cases Deaths Leukemia 621 133 Lymphomas and Reticuloendothelial Neoplasms 441 23 Central Nervous System and Miscellaneous Intracranial and Intraspinal Neoplasms (Brain Cancer) 460 78 Renal Tumors (Kidney Cancer) 121 8 Hepatic Tumors (Liver Cancer) 40 10 Malignant Bone Tumors 137 29 Sympathetic Nervous System Tumors 174 Retinoblastoma (Eye Cancer) 49 Soft-Tissue Sarcomas 199 Germ Cell, Trophoblastic and Other Gonadal Neoplasms (Reproductive Cancer) 175 Carcinomas and Other Malignant Epithelial Neoplasms (Skin Cancer) 257 Other and Unspecified Malignant Neoplasms (Cancer) 14 113 All Cancer Types 2688 394 3 Adapted from Childhood Cancer in Massachusetts 1990-1999 (2003), Massachusetts Department of Public Health. [...]... leukemia in male and female rats and malignant lymphoma in female mice (Burg, 2003; Shu, et al., 1999; Beliles and Totman, 1989) Another study found that male and female rats exposed to Perc developed kidney tumors and male and female mice exposed to Perc developed hepatocellular (liver) adenomas and carcinomas (U.S EPA, 199 8a) In addition, laboratory animal studies provide evidence that chlorinated... leukemia and other cancers Although the human evidence of the links between these chemicals and childhood cancer is weaker than it is for pesticides and solvents, the overall evidence from epidemiologic and toxicologic studies indicates that PAHs and dioxins may contribute to childhood cancer Several studies indicate that parental exposures in motor vehicle-related occupations (mechanic, gas station attendant,... to cigarette smoke, wood smoke, and vehicle exhausts are major sources of exposure to PAHs PAHs can enter water systems through discharges from industrial plants and wastewater treatment facilities The U.S EPA has found PAHs in almost half of the 1,430 national priority hazardous waste clean-up sites, and benzo (a) pyrene, benzo(b)fluoranthene, and PAHs are among the top 20 most frequently found toxic. .. malignant lymphoma in dogs and potentially in humans (U.S 25 EPA, 1994) Currently, 2,4-D is listed as a possible human carcinogen based on the study data described above, and on limited evidence of carcinogenicity in male rats During the late 1980s and early 1990s, the National Cancer Institute and the National Toxicology Program evaluated 51 pesticides for carcinogenicity and found that 24 demonstrated... lymphoma; and • other cancers in children (liver, soft-tissue sarcoma, Wilms’ tumor and carcinomas) The literature providing evidence of links between exposure to these chemical categories and various types of childhood cancer is summarized in the following tables 3) A review of supporting evidence from laboratory animal toxicology and adult human epidemiologic studies The report concludes with an analysis... majority of pesticide exposures for children occur from home, lawn, and garden use They have estimated that household applications of pesticides are 5 times greater than the per-acre application rate of pesticide-treated agricultural lands (Zahm and Ward, 1998) Children may be exposed while pesticides are being applied to a lawn or garden, or by playing on the lawn within 24 hours of application (Zahm and. .. that population for the period 1979 through 1991 Elevated rates existed particularly for leukemia and brain and CNS cancers In 1997, the NJDHSS and the Agency for Toxic Substances and Disease Registry (ATSDR) began an epidemiologic study to examine the potential exposures associated with the elevated childhood cancer rates in Dover Township 28 Table 8 Exposure to solvents and evidence of childhood cancer*... substances at these sites (ATSDR, 1996; Nadakavukaren, 2000) High dioxin exposures can be caused as a result of building fires and chemical accidents Lower levels of dioxin can be emitted from incinerator stacks and industrial sources and then carried through the air, settling on grasses eaten by grazing animals and in bodies of water where they are ingested by fish Dioxin enters the human food chain... and Wilms' tumor, Ewing’s sarcoma, neuroblastoma, and other malignancies in children was weak or inconclusive The evidence on the connections between pesticide exposure and various types of childhood cancer are summarized below, along with results of key studies Childhood cancers of concern (leukemia, brain cancer, NHL, soft-tissue sarcoma, and Hodgkin’s lymphoma) are generally the same cancers that... (sperm and eggs) of the mother or father are damaged prior to pregnancy Toxic substance exposures can cause cell damage (mutations) in the germ cells that can then be passed on to the developing embryo, causing cancer later in childhood Also, a fetus may be exposed to chemicals or pesticides during gestation Some researchers have found that substances to which pregnant women are exposed can cross the placenta . diagnosed cancers than white and Asian and Pacific Islander children. Although childhood cancer is a relatively rare disease, cancer rates increased. increases in childhood cancer rates over the last two decades and the causes of the increase, a growing body of evidence from laboratory studies and human